Development and Testing of Modules for Identification and Optimization of an Adaptive Controller

Adaptive systems consist of two main blocks: identification and optimization. The first one is responsible for determining new parameters of the changed object, the second one is to find optimal settings for the updated object. In this work, the method of applying harmonic influences is adopted. Due to the features that make it possible to better identify harmonics with a frequency of known parameters against the background of random disturbances in the object. To minimize the identification time, the method proposed by the authors is based on negative correlation was used, in which only two ordinates of one period are processed. The good efficiency of the method can be justified by the fact that the frequency of disturbances is low, in contrast to a control system with a high natural frequency. When developing an optimization block, the influence of identification error on optimality criteria was studied. To level out errors, the control principle based on the forecast of the controlled variable was applied. It allows you to obtain high-quality transient processes even from one empirically obtained object vector. In order to confirm the statements made, a calculation of the adaptive temperature control system in the section of the RF boiler of the 800 MW power unit was carried out. The calculation results convincingly prove the reliability of the proposed methodologies.